The Distribution of Wolbachia in Fig Wasps
Total Page:16
File Type:pdf, Size:1020Kb
Received 31 October 2001 Accepted 9 June 2002 Published online 2 October 2002 The distribution of Wolbachia in fig wasps: correlations with host phylogeny, ecology and population structure D. DeWayne Shoemaker1*, Carlos A. Machado2, Drude Molbo3, John H. Werren4, Donald M. Windsor5 and Edward Allen Herre5 1Department of Biological Sciences, 3149 Wood Hall, 1903 West Michigan Avenue, Western Michigan University, Kalamazoo, MI 49008-5410, USA 2Department of Genetics, Rutgers University, Nelson Biological Laboratories, 604 Allison Road, Piscataway, NJ 08854-8082, USA 3Universite de Lausanne, IE-ZEA, Batiment de Biologie, CH-1015, Lausanne, Switzerland 4Department of Biology, University of Rochester, Rochester, NY 14627, USA 5Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Republic of Panama We surveyed for the presence and identity of Wolbachia in 44 species of chalcid wasps associated with 18 species of Panamanian figs. We used existing detailed knowledge of the population structures of the host wasps, as well as the ecological and evolutionary relationships among them, to explore the relevance of each of these factors to Wolbachia prevalence and mode of transmission. Fifty-nine per cent of these wasp species have Wolbachia infections, the highest proportion reported for any group of insects. Further, neither the presence nor the frequency of Wolbachia within hosts was correlated with the population structure of pollinator hosts. Phylogenetic analyses of wsp sequence data from 70 individuals representing 22 wasp species show that neither the close phylogenetic relationship nor close ecological association among host species is consistently linked to close phylogenetic affinities of the Wolbachia associated with them. Moreover, no genetic variation was detected within any Wolbachia strain from a given host species. Thus, the spread of Wolbachia within host species exceeds the rate of horizontal transmission among species and both exceed the rate of mutation of the wsp gene in Wolbachia. The presence and, in some cases, high frequency of Wolbachia infections within highly inbred species indicate that the Wolbachia either directly increase host fitness or are frequently horizontally transferred within these wasp species. However, the paucity of cospeciation of Wolbachia and their wasp hosts indicates that Wolbachia do not persist within a given host lineage for long time-periods relative to speciation times. Keywords: Agaonidae; fig; fig wasp; horizontal transfer; Wolbachia; wsp 1. INTRODUCTION Gittenberger 1998; Wenseleers et al. 1998; Vavre et al. 1999). The proportion of insects infected by Wolbachia Wolbachia are alpha-proteobacteria that are found in a varies substantially among these more focused surveys. wide range of insects, filarial nematodes, arachnids and Thus, it appears that some insect groups, especially isopods (Bandi et al. 1998; O’Neill et al. 1992; Werren et Hymenoptera, are more prone to harbour Wolbachia than al. 1995b). Surveys of the distribution of Wolbachia across others (Wenseleers et al. 1998; Vavre et al. 1999; a broad range of arthropods have revealed a striking con- Werren & Windsor 2000). cordance in overall frequencies of infected species (17– In most insect species that have been examined in 22%) from three different geographical regions (Panama, detail, Wolbachia transmission appears to occur predomi- Britain and North America), consistent with a widespread nantly via the egg cytoplasm. Therefore, many strains are equilibrium frequency in Wolbachia infections (Werren et able to enhance their own spread within host populations al. 1995a; West et al. 1998; Werren & Windsor 2000). by a variety of mechanisms that either increase their host’s Extrapolation of these infection frequencies suggests that relative investment in daughters, or decrease the repro- Wolbachia may be among the most abundant group of ductive success of uninfected female hosts (Werren 1997). parasitic bacteria known, infecting up to five million spec- These mechanisms include cytoplasmic incompatibility, ies of insects alone (Werren et al. 1995a; West et al. 1998; thelytokous parthenogenesis, feminization of genetic Werren & Windsor 2000). In addition to these broad sur- males, male-killing, increased mating success of infected veys, there have been several taxon-specific or ecologically males via sperm competition and complete dependence on focused surveys for Wolbachia (Werren & Jaenike 1995; bacteria for egg production by the hosts (Breeuwer & Bourtzis et al. 1996; Hariri et al. 1998; Schilthuizen & Werren 1990; O’Neill et al. 1992; Beard et al. 1993; Stou- thamer & Werren 1993; Sinkins et al. 1995a,b; Wade & Chang 1995; Hurst et al. 1996; Dedeine et al. 2001; for * Author and address for correspondence: Department of Entomology, 643 Russell Laboratories, 1630 Linden Drive, University of Wisconsin, recent reviews see Werren (1997), Werren & O’Neill Madison, WI 53706-1598, USA ([email protected]). (1997) and Stouthamer et al. (1999)). Nonetheless, even Proc. R. Soc. Lond. B (2002) 269, 2257–2267 2257 2002 The Royal Society DOI 10.1098/rspb.2002.2100 2258 D. D. Shoemaker and others Wolbachia in fig wasp communities as documented transmission of Wolbachia within species or more wasp species that compete with the pollinators appears to be mostly vertical, the nearly complete lack of for oviposition sites, larger gall-forming wasps, and wasp concordance between phylogenies of Wolbachia and their parasitoids that attack the gall-formers (Boucek 1993; insect hosts demonstrates that pervasive horizontal trans- West & Herre 1994; Herre 1996; West et al. 1996). All of mission of Wolbachia among species also occurs (O’Neill these wasp species complete their life cycle in the enclosed et al. 1992; Werren et al. 1995b; Werren 1997; Zhou et al. environment of the fig syconium. Within this community 1998; Huigens et al. 2000; but see Schilthuizen & Stou- of wasps associated with any particular fig species, some thamer (1998); West et al. (1998)). species have very direct interactions with each other (e.g. More generally, the distribution of Wolbachia strains gall-formers and their parasitoids), while others have less within and among host species fundamentally depends direct interactions (e.g. the gall-formers and the upon three processes: (i) the rate of spread and the dur- pollinators). The ecological relationships among the rep- ation of maintenance of given Wolbachia infections within resentatives of these taxa are maintained across different host species; (ii) the rate of interspecific horizontal transfer host fig species and effectively represent replicate com- of Wolbachia infections across species; and (iii) the rate of munities (Wiebes 1979, 1995; Herre 1989, 1996; West & mutation of Wolbachia genes. The rate of spread within a Herre 1994; Herre et al. 1996, 1997; West et al. 1996). host species is influenced by the interaction between the Further, the phylogenetic relationships among the pollina- various effects that Wolbachia induce in their host repro- tors and the non-pollinators associated with the Panaman- duction (i.e. fitness effects of Wolbachia on their hosts, ian figs are known (Herre 1996; Herre et al. 1996; maternal transmission fidelity, etc.) and host population Machado et al. 1996, 2001; Machado 1998). Also, cali- structure. All else being equal, the rate of spread and pro- brations of fig wasp molecular clocks based on fossil evi- portion of infected individuals is expected to be lower dence indicate that the fig-pollinating wasps began to within hosts characterized by more inbred population radiate roughly 90 Myr ago and that the extant Panaman- structures (Wade & Stevens 1994). With respect to hori- ian pollinator species are very old, usually more than 3– zontal transfer of Wolbachia among different species, it is 5 Myr (Machado et al. 2001). This detailed knowledge generally expected to occur more readily among hosts that allows us to study the distribution of Wolbachia strains live in close proximity with each other or that share inti- both within and among host species in the context of vary- mately connected life cycles (e.g. parasitoids and their ing degrees of spatial proximity, variably direct linkage of hosts). life cycles and a continuum of population structures, all Importantly, the relative magnitude of genetic differ- across a known set of phylogenetic relationships. ences observed among Wolbachia strains infecting host species that are characterized by different evolutionary and 2. MATERIAL AND METHODS ecological relationships will give an indication of the most frequent scenarios of transmission among species. For (a) Collection of insect species example, if close ecological association of potential hosts All Panamanian wasps were collected on or near Barro Color- promotes Wolbachia transmission among them, we might ado Island, Panama, whereas Old World (OW) (Australia–Asia) expect that ecologically related hosts should share geneti- wasps were used to provide a phylogenetic context for the New cally similar Wolbachia strains. Further, the relative rates World (NW) species (detailed descriptions of localities are avail- of spread both within and among host species can be cali- able upon request). Fig fruits were collected late in the fruiting brated roughly against the rate of mutation of Wolbachia cycle but prior to the emergence of any wasps, returned to the genes such as wsp. For example, if the within-species laboratory, cut open and placed in Petri dishes to allow all of transmission